Peroxynitrite Oxidation of Sulfhydryls THE CYTOTOXIC POTENTIAL OF SUPEROXIDE AND NITRIC OXIDE*

Peroxynitrite anion (ONOO-) is a potent oxidant that mediates oxidation of both nonprotein and protein sulfhydryls. Endothelial cells, macrophages, and neutrophils can generate superoxide as well as nitric ox- ide, leading to the production of peroxynitrite anion in vivo. Apparent second order rate constants were 5,900 M-’*s-’ and 2,600-2,800 M-’-s“ for the reaction of peroxynitrite anion with free cysteine and the single thiol of albumin, respectively, at pH 7.4 and 37 “C. These rate constants are 3 orders of magnitude greater than the corresponding rate constants for the reaction of hydrogen peroxide with sulfhydryls at pH 7.4. Unlike hydrogen peroxide, which oxidizes thiolate anion, peroxynitrite anion reacts preferentially with the undissociated form of the thiol group. Peroxynitrite oxidizes cysteine to cystine and the bovine serum albumin thiol group to an arsenite nonreducible product, suggesting oxidation beyond sulfenic acid. Peroxynitrous acid was a less effective thiol-oxidizing agent than its anion, with oxidation presumably mediated by the de- composition products, hydroxyl radical and nitrogen dioxide. The reactive peroxynitrite anion may exert cytotoxic effects in part by oxidizing tissue sulfhydryls.

mediator of cell injury (11). Superoxide is implicated as a toxic species in free radical-mediated cytotoxicity when superoxide dismutase limits cell injury. For example, augmentation of intracellular superoxide dismutase prevents oxygeninduced damage in endothelial cells (12). However, 0; chemical reactivity is limited, compared to other free radicals (13). Whereas 0; can react with critical cellular targets to directly exert toxicity (14), enzymatic and spontaneous dismutation can be effective competing reactions that may limit direct toxic effects of 0 2 (14, 15). While other partially reduced oxygen species such as hydrogen peroxide (H202) and hydroxyl radical ('OH) are cytotoxic, hydroxyl radical is frequently proposed as an ultimate cytotoxic agent due to its high reactivity (16). Hydroxyl radical can be generated through an 0;-driven Fenton reaction, in which 0, plays the role of a reducing agent (17).
H20, + Fez+ + 'OH + OH-+ Fe3+ (2) Our current understanding of oxygen-mediated toxicity via the Haber-Weiss and Fenton reactions does not always give a consistent explanation for protective effects lent by superoxide dismutase. It has been suggested that 0; serves as a reductant for Fe3+ (Equation 1) and that superoxide dismutase blunts 'OH production by preventing Fez+ reduction of HzOz (17). Other investigators have shown that ' OH formation can be supported by other biological reductants of Fe3+. Ascorbate or glutathione are present at much higher concentrations than 0; in cells and can alternatively serve to reduce Fe3+ under physiological conditions (18). Thus, other reactions than those leading to .OH formation may be important in understanding mechanisms of 0;-dependent toxicity and the protective role of superoxide dismutase.
An important biological reaction of 0; may be with endothelial-derived 'NO, a free radical species. Macrophages and neutrophils are also capable of production and release of both 0; and 'NO (19,20). The reaction between these two radical species is diffusion-limited in gaseous phase and is extremely rapid in aqueous phase, yielding peroxynitrite (ONOO-; Ref. 21). Potentially toxic levels of peroxynitrite can be achieved in tissues under conditions when 'NO and 0; production are stimulated, because a 100-fold increase in the rate of peroxynitrite formation should occur for every 10-fold increase in ' NO and 0; concentration.
Peroxynitrite has a pK, of 6.8 at 37 "C. Peroxynitrous acid is unstable with a half-life of under 1.0 s (22), decomposing to give a species with hydroxyl radical-like reactivity, according to the following reaction.
Peroxynitrous acid was revealed as a highly reactive species generating a strong oxidant capable of oxidizing deoxyribose and dimethyl sulfoxide with high yields at acidic p H (22). Nitrogen dioxide is the other product of peroxynitrite decomposition and is also a strong oxidant with significant cytotoxic potential (23-25).
We investigated the reactions of peroxynitrite towards protein and nonprotein sulfhydryl groups and compared this with H202-mediated thiol oxidation. Sulfhydryls are common targets for free radicals, with sulfhydryl oxidation being a key mechanism of free radical-mediated toxicity at the molecular level (26). Thiols are critical to the active site of many enzymes and for maintaining the native conformation of proteins (27). Herein, we report that peroxynitrite anion oxidizes sulfhydryls about IO3 times faster than does H202 a t 37 "C and p H 7.4 and propose that this reaction may be an important mechanism of oxygen radical-mediated toxicity.
Albumin Preparation-Bovine serum albumin was dissolved to 1.52 mM in 0.01 M potassium phosphate, pH 7.4, and dialyzed overnight in 10 mM 2-mercaptoethanol in the same buffer to reduce sulfhydryls oxidized during purification and storage. After 36-h dialysis with six changes of 0.01 M potassium phosphate, BSA typically had a sulfhydryl to albumin molar ratio of 0.6-0.7, in accordance with previous analyses of fresh plasma albumin (28). The thiol group of mercaptoethanol-treated BSA was chemically modified in three ways; 1.4 mM BSA (SH/BSA = 0.7) was incubated for 30 min a t 18 "C with either 10.6 mM H,Op, 6.4 mM NEM, or 1.03 mM mercuric chloride (HgC12). Excess H20, or NEM was removed by chromatography of BSA on Sephadex G-25. NEM or HgC1, treatment left no residual sulfhydryls whereas H202 treatment gave a final SH/BSA ratio of 0.16. Protein was quantified with a bicinchoninic acid assay (29).
quenched flow reactor (30). Solutions of (a) 0.6 M NaN02 and (b) 0.6 Peroxynitrite Synthesis-Peroxynitrite was synthesized in a M HCl plus 0.7 M H,02 were pumped a t 26 ml/min into a T junction and mixed in a 3-mm-diameter by 2.5-cm glass tube. The acidcatalyzed reaction of nitrous acid with H202 to form peroxynitrous acid was quenched by pumping 1.5 M NaOH at the same rate into a second T junction at the end of the glass tubing. Excess H,O, was removed by passage over a 1 X 5-cm column filled with 4 g of granular Mn02. The solution was frozen a t -20 "C for as long as a week. Peroxynitrite forms a yellow top layer due to freeze fractionation, which was retained for further studies. This top layer typically contained 170-220 mM peroxynitrite as determined by absorbance a t 302 nm (c:roz = 1,670 M".cm"; Ref. 31).
Sulfur Chemistry-Sulfhydryls were quantified using DTNB at 412 nm (e1', = 1.36 X IO4 ~" . c m " ; Ref. 32). A 10 mM DTNB stock solution was prepared, and reactions were performed in 0.05 M potassium phosphate, 2 mM EDTA, pH 9.0, for 20 min at room temperature. In some experiments, pCMPS was used for determination of sulfhydryl content; pCMPS was prepared as a 2 mM stock solution, and sulfhydryl content was detected in 0.05 M Tris-HCI, pH 7.4, for 10 min a t room temperature (ez4" = 7,470 M-l.cm" for cysteine). Final DTNB and pCMPS concentrations used were 0.20 mM, always in excess over assay sulfhydryl concentration.
Sulfur-reducing agents, including sodium borohydride, sodium arsenite, potassium cyanide, and DTT were used for reductive recovery of oxidized thiol groups. Oxidized DTT was measured by its absorbance at 283 nm calculated from the ratio between the "recovered" thiols in the presence of the reductants versus the initial amount of oxidized thiols.
In Equation 4, So and S, represent initial and final thiol concentrations, and S, is the thiol concentration after addition of reductant to oxidized samples. Kinetic Experiments-Peroxynitrite decomposition in the presence of sulfhydryls was studied by stopped flow absorbance spectroscopy a t 302 nm (Hi-Tech Instruments). The kinetics of peroxynitrite decomposition were fitted to a pseudo-first-order reaction equation by nonlinear regression. The nitrogen-containing species derived from peroxynitrite reaction (ie. nitrite and nitrate) absorb at 340 nm and did not interfere with the 302 nm measurements. A typical run consisted of 500 points collected over 0.1-10 s, so that at least 99% of peroxynitrite disappeared.
Thiol oxidation by H20, was determined by taking timed aliquots for sulfhydryl content analysis.

Equation 5 represents the reaction between He02 and sulfhydryls,
where Ho and So represent the initial concentration of H202 and sulfhydryl, respectively, S is sulfhydryl consumed after time t, n is moles of sulfhydryl oxidized per mole of peroxide and k A , y is the apparent second order rate constant of the reaction. The value of n was assumed to be 2 for the reaction between cysteine and H20, (i.e. formation of cystine) and 1 for the reaction between BSA-SH and Hs02 (i.e. formation of BSA-sulfenic acid as the first intermediate).
UV Difference Spectroscopy-Difference spectra were recorded on a Gilford response spectrophotometer to determine the pK, of BSA thiol ionization. Difference spectra were recorded by subtraction, with a BSA reference solution in 0.1 M potassium phosphate, pH 5.0. To diminish the influence of other species when titrating BSA thiolate anion, difference spectra of both mercapto-reduced and H,02-oxidized BSA were recorded a t different pH. Absorbance at 237 nm was used to calculate the pK, of the BSA thiol group.

RESULTS
Peroxynitrite Reaction with Sulfhydryls-In the presence of a 20-100-fold excess of cysteine, the disappearance of peroxynitrite followed pseudo-first-order kinetics when monitored a t 302 nm by stopped flow spectroscopy (Fig. 1). Under these conditions, the spontaneous decomposition of peroxynitrite appeared as a small non-zero intercept (Fig. lb). Cysteine and cystine do not absorb significantly at 302 nm, thus not interfering with peroxynitrite quantitation. The pH dependence of the second-order rate constant ( Fig. 2) suggests that cysteine reacts rapidly with peroxynitrite anion. The reaction can be described by where k2' is the apparent rate constant at a given pH, k1 is the second-order rate constant of cysteine reaction with peroxynitrite anion, K,, is the dissociation constant of peroxynitrite and Ka2 is the dissociation constant of cysteine to the thiolate anion. The best fit of the data in Fig. 2 gives kS = 5,900 f 160 M".s" with pK,, = 6.6 and pK,, = 8.3. The pK, of peroxynitrite at 37 "C is 6.8.2 The pK, of 7.5 at 37 "C reported previously for peroxynitrite (22) was in error, due to instrument problems. The apparent pK,, of 8.3 corresponds to the ionization of cysteine.
The overall apparent rate constant for BSA reacting with peroxynitrite was 5,020 f 480 M" .s" (mean f S.D., n = 7)  The rate constant of BSA sulfhydryl reaction with peroxynitrite was also estimated by reaction of BSA with limiting amounts of peroxynitrite and assessment of residual BSA sulfhydryl content (Fig. 3). Under these conditions, the rate of peroxynitrite decomposition may be written as where [SH] is the concentration of the sulfhydryl compound, kSP is the second-order rate constant between peroxynitrite and SH, and k, is the sum of all peroxynitrite decomposition processes, which include spontaneous decomposition and other reactions with BSA. Thus, kx implicitly depends upon BSA concentration. Because only a small amount of SH is consumed in this reaction, the integral of Equation 7 can be approximated by where [SHIo and [ONOO-Io are the initial concentrations of SH and peroxynitrite. The rate of SH oxidation is expressed as follows. amino acid residues of BSA, including sulfhydryls. When sulfhydryls were selectively removed by prior treatment with H202, NEM, or HgC12, the rates of reaction were 2,400 +_ 240 ( n = 6), 2,300 f 95 ( n = 6), and 2,600 f 290 (n = 3) M".s-', respectively. The difference between the first rate constant and the average of the latter three suggests that the free sulfhydryl of BSA reacts with peroxynitrite at a rate of The left-hand expression can be determined as the slope from Fig. 3, which was -1,490 M" for 0.340 mM BSA sulfhydryl. The term k, was taken to be 0.911 f 0.059 by averaging the three estimates of 0.898 f 0.089, 0.861 f 0.036, and 0.977 f 0.107 s-l from stopped flow measurements of peroxynitrite disappearance at 302 nm using the same concentration of BSA after sulfhydryl reaction with H202, NEM, or HgCl,. Substituting these values and rearranging Equation 12 to solve ksP gave a rate constant of 2,800 f 180 M".s". Thus, two separate methods for approximating the rate constant of peroxynitrite reaction with the BSA sulfhydryl gives estimates that are about 50% of values observed with free cysteine.
Effects of Metal Chelators-Desferrioxamine was a concentration-dependent inhibitor of dimethyl sulfoxide and deoxyribose oxidation by peroxynitrous acid (22). In contrast, the metal chelator DTPA had no effect on the extent of sulfhydryl oxidation by peroxynitrite (Fig. 3). Desferrioxamine tested in a broad concentration range (25-500 PM) at pH 7.3 and 8.0 also had no significant influence on sulfhydryl oxidation (data not shown).
Oxidative Yield. and pH-Sulfhydryl oxidation of BSA or cysteine by peroxynitrite occurred at acidic pH, but was greater at alkaline pH, having a 50% oxidation yield at pH 6.84 f 0.09 (Fig. 4,O) for BSA and pH 7.02 f 0.06 (Fig. 4, A) for cysteine at 37 "C. Under our experimental conditions, the extent of sulfhydryl oxidation was about 3-4 times greater for cysteine than for BSA. Cysteine, in a 10-fold excess over peroxynitrite (2.5-0.25 mM) at pH 9.0 (thus making spontaneous decomposition insignificant), was oxidized in a ratio of 2.2 f 0.1 (mean f S.D., n = 3) mol of cysteine/mol of peroxynitrite.
Reaction with H202-The reaction of H202 with BSA-SH or cysteine obeyed a second-order rate law ( r > 0.99), giving apparent rate constants of 1.14 k 0.03 M".s" for BSA-SH and 4.69 f 0.06 M". s-' for cysteine a t 37 "C and pH 7.4 (Fig.  5 ) . Reaction rates were greater at alkaline pH (Fig. 6). Fitting the results to the Henderson-Hasselbalch equation gave pK, values of 8.00 k 0.07 for the reaction of H202 with BSA-SH and 8.04 f 0.07 for the reaction of H202 with cysteine. Since H202 has a pK, of 11.8 (34), the pH dependence of sulfhydryl oxidation was more closely related to the dissociation of the thiol group to the thiolate anion (35).  The first reaction between thiolate and H20z leads to the formation of sulfenic acid. The overall rate of reaction was thus given by where Kd represents the dissociation constant of the thiol group and kHS is the rate constant of RS-reacting with H202. The upper limit of k;ls for both curves in Fig. 6 tends to the rate constant (kHs) with values of 17.1 k 0.5 and 5.6 f 0.2 M".s" at 37 "C for the reaction of HzOz with cysteine and BSA-SH, respectively. In the neutral to alkaline pH range, Peroxynitrite and Sulfhydryl Oxidation cysteine reacted 2-4 times faster with Hz02 compared with the reaction of BSA-SH with Hz02. During cysteine oxidation to cystine, it has been assumed by us and others (35) that the rate-limiting reaction is the formation of the cysteine sulfenic acid intermediate.

UV
Difference Spectroscopy and pK, Determination of the BSA Sulfhydryl-Ultraviolet difference spectroscopic analysis of BSA at pH 8.0 uersus pH 5.0 absorbed at 230-240 nm, indicative of thiolate anion (Fig. 7). At pH < 7.0, oxidized BSA absorbed more strongly than native BSA, whereas at pH > 9.0 the absolute value of the spectroscopic difference decreases. Thus, other substituent groups or conformational changes of BSA are also involved in 230-240 nm absorbance. We determined an 6237 = 5,340 M" cm" for the thiolate anion of BSA, similar to previously reported extinction coefficients for other protein and nonprotein thiolate anion species (36).
Titration of BSA above pH 7.5 increased absorbance of the thiolate anion, maximal at pH 8.0-8.5 (Fig. 7). The 50% increase of absorbance at pH 7.86, close to the pK, of the reaction of BSA-SH with H202, supports the role of BSA thiolate anion as the sulfur species reactive with ONOO-.
Thiol Oxidation State-Arsenite, cyanide, and dithiothreitol-mediated reduction of oxidized thiols permitted determination of the reversibility of oxidation processes and aided in identification of thiol oxidation states (Fig. 8). Arsenite reduces sulfenic acid (RSOH) to sulfhydryl (RSH) but does not reduce disulfides (RSSR Ref. 37). Cyanide and DTT reduce disulfides to thiols with different reaction stoichiometries. RSSR + CN-"* RS-+ RSCN (16) RSSR + (CHOH"CH2"SH)z "* 2RSH (17) One mole of cyanide and DTT recovers 1 and 2 mol of RSH from disulfide, respectively. Under our assay conditions, NaBH4-reduced BSA had a sulfhydryl to BSA ratio of 1.0 in control samples, suggesting total reduction of the single cysteine sulfur of BSA not participating in intramolecular disulfide formation. Higher NaBH4 concentrations (-25-fold greater) were used by others for reduction of constitutive disulfides of BSA (38).
Peroxynitrite-mediated BSA-SH oxidation gave a product resistant to arsenite reduction, whereas NaBH4 reduction of oxidized BSA thiol was 75% (Table I). In contrast, 52% of H2O2-mediated BSA-SH oxidation was reduced by arsenite, suggesting a significant amount of sulfenic acid formation (Table I). Sodium borohydride similarly reduced 90% of the HzOz-oxidized BSA sulfhydryls (Table I). Electrophoretic analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed that neither peroxynitrite nor HzO, resulted in disulfide-mediated dimerization or peptide cleavage of BSA at pH 7.4 (data not shown). Intramolecular disulfide formation is not possible since BSA has only one thiol. Peroxynitrite-dependent cysteine oxidation was 90% cyanide-and DTT-reducible, with arsenite having no effect. Hydrogen peroxide-mediated cysteine oxidation was 54% cyanide-reducible and 48% DTT-reducible; arsenite had no significant effect (Tables I1 and 111). Cysteine thiols were measured using pCMPS, to avoid artifact from cyanide dependent-DTNB reduction. Similar results were obtained when DTNB was used in place of pCMPS for quantitating arsenite reduction of cysteine oxidation products.

DISCUSSION
Peroxynitrite mediated the oxidation of both nonprotein and protein sulfhydryls. The higher yields of sulfhydryl oxidation at alkaline pH indicate that peroxynitrite anion, rather than peroxynitrous acid, was the primary oxidizing species. In the acidic pH range, some oxidation still occurred, most

Arsenite and borohydride reduction of hydrogen peroxide and peroxynitrite-oxidized BSA sulfhydryls
Bovine serum albumin (0.48 mM) was incubated with 0.20 mM Hz02 or 0.25 mM peroxynitrite (final concentration) for 20 min at 37 "C in 2.0 ml of 0.1 M potassium phosphate, pH 7.4. Then, 37 mM sodium arsenite or 26 mM sodium borohydride was added to aliquots of reaction mixtures and incubated for 45 min at 37 "C. Samples were immediately assayed for sulfhydryl content. Data represents f +. S.D., n = 3. * Significantly different from the "no reductants" condition in the presence of the oxidant (H202 or ONOO-). e Significantly different from the "no reductants" condition in the absence of the oxidant (Hz02 or ONOO-).

SH/BSA
Significantly different from matched treatment groups without the oxidant (HZ02 or ONOO-).
Significance of difference was determined by ANOVA and analysis by Duncan's multiple range test ( p < 0.05).

Cyanide and arsenite reduction of peroxynitrite and hydrogen peroxide-oxidized cysteine
Cysteine (0.36 mM) was incubated with 0.96 mM peroxynitrite or 0.90 mM H202 for 15 min at 37 "C in 0.05 M Tris-HC1, pH 7.4. Then, 23.5 mM potassium cyanide or 23.5 mM sodium arsenite was added to aliquots of the reaction mixtures incubated for 30 min at 37 "C in 0.05 M Tris-HC1, pH 7.4. Samples were immediately assayed for sulfhydryl content with pCMPS. Data are the mean of two different analyses from a representative experiment.  Cysteine (1.3 mM) was incubated with either 2.0 mM peroxynitrite or 2.4 mM Hz02 in potassium phosphate, pH 7.4, for 10 min at 37 "C. Aliquots were removed, and 7 units/ml catalase was added to eliminate the remaining H202. Then, 5 mM DTT was immediately added, and reaction mixtures were incubated for 15 min at 37 "C. Disulfide content was determined from absorbance at 283 nm. likely due to decomposition of peroxynitrous acid. This process would produce strong oxidants with reactivity like OH' and nitrogen dioxide that can attack sulfhydryls. The pH dependence of the reaction rates of peroxynitrite anion towards thiols (Fig. 2) indicates that peroxynitrite anion reacts preferentially with the undissociated form of the thiol group (SH). At pH values above the pK, of the thiol, oxidation yields are still high (Fig. 4) because the spontaneous decomposition of peroxynitrite is slower than at acidic pH. Thiolate anion species are usually more susceptible to oxidation (37), but this would not be expected for peroxynitrite anion due to electrostatic repulsion between the two negatively charged species. When H202 was the oxidant, the pH dependence of sulfhydryl oxidation followed thiolate anion dissociation (H202 pK. = 10.8). The pK, of the cysteine sulfhydryl at 25 "C is 8.3-8.5 (39), and the heat of ionization of S H groups is about 7 cal/ mol. Thus, the pK, should decrease by 0.2 when temperature increases from 25 to 37 "C. Thus, a pK, for cysteine of 8.10-8.30 at 37 "C is close to the pK, value of 8.04 reported herein, confirming that the thiolate anion was the reactive sulfur species in Hz02-mediated oxidation reactions.
The sulfhydryl of BSA had a pK, of 7.86-8.00, according to both spectroscopic and oxidation kinetics data. The pK, of protein sulfhydryls may differ significantly from the pK. of free cysteine sulfhydryls, depending upon the sulfhydryl microenvironment (27). Nevertheless, the pK, of BSA-SH was similar to cysteine. The reactivity of peroxynitrite and H202 towards BSA-SH was probably less than that of cysteine due to steric restrictions.
Peroxynitrite is more reactive with sulfhydryls than H202. According to the effective second-order rate constants, we observed that peroxynitrite reacted about 2,600 times faster towards BSA-SH and about 1,200 times faster towards cysteine than did H202 at 37 "C, p H 7.4 ( Figs. 1 and 5). The rate constant determined for the reaction of H202 with cysteine agrees with that previously reported, ranging between 10 and Sulfhydryl oxidation analysis after reaction of HzOz and peroxynitrite showed different products. Electrophoretic analysis showed that intermolecular disulfide bond formation by BSA after peroxynitrite or H202 oxidation did not occur. Steric restrictions due to the size of the albumin molecule or the location of the single albumin thiol preclude disulfide formation, only occurring with denaturing conditions (40). Sterically isolated S H groups can be oxidized beyond the disulfide, with a t least 52% of the BSA-SH oxidized to sulfenic acid (RSOH) by H202, and the remainder of the sulfhydryl groups further oxidized to possibly sulfinic (RSO;) or sulfonic (RSO;) acid (Table I). In agreement with our observations, the BSA-SH is oxidized by iodine or thiocyanogen to sulfenyl iodide or sulfenyl thiocyanate, respectively, oxidation states equivalent to sulfenic acid (37). Under stronger H202-mediated oxidizing conditions, no sulfenic acid was detectable (41). Peroxynitrite did not oxidize BSA-SH to sulfenic acid, but a high proportion of peroxynitrite-oxidized sulfhydryls were borohydride-reducible. Apart from the eventual formation of sulfinic or sulfonic acid after peroxynitrite oxidation, other products of the type RSNO, may have been formed, many of which are unstable in H20. For example, the oxidation of glutathione by tetranitromethane (C(N02),) formed a sulfinic acid derivative, with sulfenyl nitrate formed (RSNO,) as a transient intermediate (42). Such an intermediate could be stabilized in an appropriate protein microenvironment.
Cysteine was oxidized with an excess of either H202 or peroxynitrite to predominantly the disulfide form, with no formation of sulfenic acid. The main product of H202 oxidation of cysteine beyond the disulfide is cysteic (sulfonic) acid (37). Peroxynitrite gave a >90% yield of disulfide from cys-teine (Table 111), similar to that observed for the oxidation of thiols by . NO and NOz (25).
Transition metals did not have a significant role in peroxynitrite-mediated sulfhydryl oxidation, since neither DTPA nor desferrioxamine modified oxidation yields (Fig. 3). Desferrioxamine, but not DTPA, is a strong inhibitor of peroxynitrous acid-dependent oxidation of deoxyribose and dimethyl sulfoxide in a manner unrelated to its iron-chelating ability (22). No significant effect of desferrioxamine was detected in our system, supporting the concept that desferrioxamine reacts preferentially with peroxynitrous acid rather than with peroxynitrite anion.
Peroxynitrite-mediated oxidation of sulfhydryls has significant cytotoxic potential. First, oxidation of low molecular weight sulfhydryls (i.e. cysteine and glutathione) leads to depletion of one of the most important intra-and extracellular scavenging mechanisms serving as a defense against free radical-mediated damage. This increases the possibility of alteration and destruction of critical macromolecules such as DNA, enzymes, structural proteins, structural polysaccharides, and membrane phospholipids. Second, oxidation of protein sulfhydryls and thiol-containing cofactors (i.e. coenzyme A, lipoic acid, thioredoxin, and phosphopantetheine) can perturb integrated metabolic pathways and membranelinked functions that participate in essential metabolic and biosynthetic processes. The glutathione-glutathione reductase system should be able to recover, at least partially, peroxynitrite-derived disulfides (i.e. cystine) but not higher sulfur oxidation states derivatives formed from peroxynitrite and Hz02 reaction with isolated protein sulfhydryls. Peroxynitrite-mediated oxidation of sulfhydryls does not require dissociation of the thiol group. Peroxynitrite will react with sulfhydryls having a high pK,, such as glutathione (pK, = 9.2), considered resistant to autooxidation compared with other sulfhydryls because less than 1.6% exists as thiolate anion at physiological pH.
Peroxynitrite can either react with sulfhydryl groups or become protonated to form peroxynitrous acid, which decays rapidly and gives rise to an oxidant species having a reactivity similar to OH' (Fig. 8). In addition, NOz is formed as a decay product, which is capable of initiating fatty acid peroxidation (23), nitrosylation of aromatic amino acids (24), and sulfhydryl oxidation (25). Thus, both peroxynitrite anion and peroxynitrous acid are potentially toxic species which have differing target molecule selectivities.
Pathophysiological processes such as ischemia-reperfusion, acute inflammation, sepsis, or excessive activation of the Nmethyl-D-aspartic acid receptor in the brain can cause the influx of Ca'+ into cells, enhancing the simultaneous production of 0; and 'NO. Calcium is postulated to participate in stimulation of 0; production by either xanthine oxidase (10) or mitochondria (43) and may increase 0; production by endothelium (44). Nitric oxide biosynthesis is catalyzed by a Caz+-dependent arginine oxidase activity (4). It has been estimated that the intraluminal rate of 'NO production in small blood vessels could reach values as high as 8 p M . min" (22). Once simultaneously produced, 0; and 'NO can react with each other, rapidly forming peroxynitrite. itloreover, 0; and 'NO could be released separately from different cell types and form peroxynitrite intravascularly or in extracellular compartments. Superoxide dismutase would effectively prevent the reaction between 0; and 'NO by scavenging 0; at almost diffusion-controlled rates.
The bolus concentrations of peroxynitrite used in the present study are relatively high, compared to steady-state concentrations of peroxynitrite that might occur in uiuo. Nevertheless, if one considers dosage in terms of time x concentrution, bolus addition of 100 p~ peroxynitrite (decaying only by proton-catalyzed decomposition) would be equivalent to a physiologically relevant steady-state concentration of 2.8 pM for 1 min.
Superoxide-derived peroxynitrite is highly reactive toward sulfhydryls, unlike superoxide-derived H20z. These differential reactivities of peroxynitrite and H2OZ may well extend to other biochemical targets and provide new understanding of mechanisms of 0;-mediated toxicity.